Predicting the Influence of Additional Training Data on Classification Performance for Imbalanced Data
It is desirable to predict the influence of additional training data on classification performance because the generation of samples is often costly. Current methods can only predict performance as measured by accuracy, which is not suitable if one class is much rarer than another. We propose an approach which is able to also predict other measures such as G-mean and F-measure, which are used in cases of imbalanced data. We show that our method leads to more correct decisions whether to generate more training samples or not using a highly imbalanced real-world dataset of scanning electron microscopy images of nanoparticles.
KeywordsTraining Sample Classification Performance Classifier Performance Misclassification Rate Class Imbalance
- 2.Kockentiedt, S., Tönnies, K., Gierke, E., Dziurowitz, N., Thim, C., Plitzko, S.: Automatic detection and recognition of engineered nanoparticles in SEM images. In: VMV 2012: Vision, Modeling & Visualization, pp. 23–30. Eurographics Association (2012)Google Scholar
- 3.Kohavi, R., Wolpert, D.H.: Bias plus variance decomposition for zero-one loss functions. In: Proceedings of the 13th International Conference on Machine Learning, pp. 275–283 (1996)Google Scholar
- 6.Smith, J.E., Tahir, M.A., Sannen, D., Van Brussel, H.: Making early predictions of the accuracy of machine learning classifiers. In: Sayed-Mouchaweh, M., Lughofer, E. (eds.) Learning in Non-stationary Environments, Chap. 6, pp. 125–151. Springer, New York (2012)Google Scholar
- 8.Webb, G.I., Conilione, P.: Estimating bias and variance from data. Technical report, Monash University, Melbourne (2003). http://www.csse.monash.edu/~webb/Files/WebbConilione06.pdf